Deep Clouds Research Articles

Retraction Note: Cyber attack detection in monitoring on optoelectronics devices using deep learning model and cloud computing network

Retraction Note: Cyber attack detection in monitoring on optoelectronics devices using deep learning model and cloud computing network

IDENTIFEYE: Sketch-Based Facial Analysis in Forensics

Traditional forensic methods, like hand-drawn sketches, often fall short in quickly identifying suspects, making the process both slow and resource-intensive. This paper introduces an innovative application designed to streamline this task. Through a simple drag-and-drop interface, users can upload sketches and match them with entries in police databases, significantly boosting the speed and accuracy of suspect recognition. By integrating deep learning and cloud technology, this tool builds on existing algorithms to deliver higher precision and ease of use. Earlier methods have had trouble aligning sketches with photographs, especially when orientations differ. Our application overcomes these issues, merging the art of sketching with modern technology to enhance forensic work. Testing reveals promising accuracy, though ongoing improvements are needed to fully meet the demands of law enforcement. This research aims to offer a real-world solution, tackling past limitations and making suspect identification faster and more efficient for investigators. Keywords: Forensic identification, Suspect sketching, Deep learning, Recognition software, Law enforcement tools, Cloud technology.

Research and application on deep learning-based point cloud completion for marine structures with point coordinate fusion and coordinate-supervised point cloud generator

Research and application on deep learning-based point cloud completion for marine structures with point coordinate fusion and coordinate-supervised point cloud generator

AN EFFICIENT FRACTAL CARDIO DISEASES ANALYSIS USING OPTIMIZED DEEP LEARNING MODEL IN CLOUD OF THING CONTINUUM ARCHITECTURE

Exercise has long been known to improve cardiovascular health, energy metabolism, and well-being. However, myocardial cell responses to exercise are complex and multifaceted due to their molecular pathways. To understand cardiac physiology and path physiology, one must understand these pathways, including energy autophagy. In recent years, deep learning techniques, IoT devices, and cloud computing infrastructure have enabled real-time, large-scale biological data analysis. The objective of this work is to extract and analyze autophagy properties in exercise-induced cardiac cells in a cloud-IoT context using deep learning, more especially an autoencoder. The Shanghai University of Sport Ethics Committee for Science Research gave its approval for the data collection, which involved 150 male Sprague–Dawley (SD) rats that were eight weeks old and in good health. The [Formula: see text]-score normalization method was used to standardize the data. Fractal optimization methods could be applied to these algorithms. For example, fractal-inspired optimization techniques might be used to analyze deep learning with Autoencoder, the autography energy of exercise myocardial cells within a cloud-IoT. To capture the intricate myocardial energy autophagy during exercise, we introduced the DMO-GCNN-Autoencoder, a Dwarf Mongoose Optimized Graph Convolutional Neural Network. The results showed that the proposed network’s performance matches that of the existing methods.

A domain adaptation framework for cross-modality SAR 3D reconstruction point clouds segmentation utilizing LiDAR data

Synthetic aperture radar (SAR) 3D point clouds reconstruction can eliminate the problems of layover in 2D SAR image projections, the recognition of the reconstructed point cloud can significantly enhance target identification and information extraction. Current SAR 3D point cloud segmentation methods, based on traditional machine learning clustering approaches, suffer from poor automation and accuracy. However, the absence of publicly labeled SAR datasets impedes the progress of deep learning-based SAR 3D point cloud segmentation methods. To tackle the aforementioned challenges, we introduce, for the first time, an alternative training approach for SAR 3D point cloud segmentation using LiDAR annotated data, which offers a more abundant sample pool, thus alleviating the lack of training sets. Nevertheless, a segmentation model trained on LiDAR point clouds exhibits a significant decline in performance when directly applied to SAR 3D reconstructed point clouds due to the cross-domain discrepancy. This research presents a pioneering domain adaptation 3D semantic segmentation framework to implement cross-modal learning for SAR point clouds. In our scheme, a simple yet effective technique is developed to achieve segmentation of SAR double-bounce scattering regions called SARDBS-Mix, which employs a mixing strategy, to capture the distinctive reflection characteristics of SAR data. Furthermore, we implement approaches including center alignment and normalization, local augmentation, and weighted cross entropy to mitigate LiDAR and SAR domain gap and class imbalances. The experimental results validate the feasibility and effectiveness of the proposed method for SAR 3D reconstructed point cloud segmentation.

Digital reconstruction of railway steep slope from UAV+TLS using geometric transformer

Accurate representation of railway slopes, especially those that are steep, is vital for real-time risk perception. Also, temporary structures also present certain safety hazards due to lack of monitoring. Traditional point cloud modeling, employing Unmanned Aerial Vehicle (UAV) or Terrestrial Laser Scanning (TLS), often struggles to simultaneously account for the precision of both surface and overhead models, leading to considerable model distortion, roughness, and deviation. Addressing these issues, A new 3D point cloud modeling algorithm for railway slopes based on a geometric transformer is presented in this paper. This involves an innovative rough point cloud denoising technique leveraging adaptive segmentation, multi-scale denoising, and deep learning point cloud registration. Our approach significantly enhances UAV point cloud accuracy and supplements missing portions of the TLS point cloud dataset occluded by objects block, using data from the UAV point cloud set. An experimental study shows that the score-based denoising algorithm improves precision from 37.44 mm to 8.11 mm for a UAV 3D point cloud. Further, by registering the UAV and TLS point cloud sets using the Geometric Transformer algorithm, the precision of the 3D point cloud was further augmented to 5.11 mm, representing a sevenfold enhancement over the initial UAV point cloud accuracy prior to denoising. Consequently, a high-fidelity 3D point cloud model of steep railway slopes has been created.

Large Eddy Simulation of Vertical Structure and Size Density of Deep Layer Clouds

Large Eddy Simulation of Vertical Structure and Size Density of Deep Layer Clouds

Truth discovery for mobile workers in edge-assisted mobile crowdsensing

The proliferation of mobile phones has led to the rise of mobile crowdsensing systems. However, many of these systems rely on the deep cloud, which can be complex and challenging to scale. To improve the performance of crowdsensing at the edge cloud, truth-discovery methods are commonly employed. These methods typically involve updating either the truth or the weight associated with a user’s task. While some edge cloud-based crowdsensing systems exist, they do not provide incentives to users based on their experience. In this report, we present a new approach to truth discovery and incentive-giving that considers both the user’s experience and the accuracy of their submitted data. Our modified truth-discovery algorithm updates both the weight and truth concurrently, with greater incentives offered to users who have completed more tasks and whose submitted data is close to the estimated truth. We have run simulations to show how well our suggested strategy works to enhance the incentive system for experienced users.

Four-of-a-kind? Comprehensive atmospheric characterisation of the HR 8799 planets with VLTI/GRAVITY

With four companions at separations from 16 to 71 au, HR 8799 is a unique target for direct imaging, presenting an opportunity for a comparative study of exoplanets with a shared formation history. Combining new VLTI/GRAVITY observations obtained within the ExoGRAVITY program with archival data, we performed a systematic atmospheric characterisation across all four planets. We explored different levels of model flexibility to understand the temperature structure, chemistry, and clouds of each planet using both petitRADTRANS atmospheric retrievals and fits to self-consistent radiative–convective equilibrium models. Using Bayesian model averaging to combine multiple retrievals (a total of 89 across all four planets), we find that the HR 8799 planets are highly enriched in metals, with [M/H] ≳1, and have stellar to superstellar atmospheric C/O ratios. The C/O ratio increases with increasing separation from 0.55−0.10+0.12 for d to 0.78−0.04+0.03 for b, with the exception of the innermost planet, which has a C/O ratio of 0.87 ± 0.03. Such high metallicities are unexpected for these massive planets, and challenge planet-formation models. By retrieving a quench pressure and using a disequilibrium chemistry model, we derive vertical mixing strengths compatible with predictions for high-metallicity, self-luminous atmospheres. Bayesian evidence comparisons strongly favour the presence of HCN in HR 8799 c and e, as well as CH4 in HR 8799 c, with detections at > 5σ confidence. All of the planets are cloudy, with no evidence of patchiness. The clouds of c, d, and e are best fit by silicate clouds lying above a deep iron cloud layer, while the clouds of the cooler HR 8799 b are more likely composed of Na2S. With well-defined atmospheric properties, future exploration of this system is well positioned to unveil further details of these planets, extending our understanding of the composition, structure, and formation history of these siblings.

PointStaClu: A Deep Point Cloud Clustering Method Based on Stable Cluster Discrimination

Potential inconsistencies between the goals of unsupervised representation learning and clustering within multi-stage deep clustering can diminish the effectiveness of these techniques. However, because the goal of unsupervised representation learning is inherently flexible and can be tailored to clustering, we introduce PointStaClu, a novel single-stage point cloud clustering method. This method employs stable cluster discrimination (StaClu) to tackle the inherent instability present in single-stage deep clustering training. It achieves this by constraining the gradient descent updates for negative instances within the cross-entropy loss function, and by updating the cluster centers using the same loss function. Furthermore, we integrate entropy constraints to regulate the distribution entropy of the dataset, thereby enhancing the cluster allocation. Our framework simplifies the process, employing a single loss function and an encoder for deep point cloud clustering. Extensive experiments on the ModelNet40 and ShapeNet dataset demonstrate that PointStaClu significantly narrows the performance gap between unsupervised point cloud clustering and supervised point cloud classification, presenting a novel approach to point cloud classification tasks.

Atmospherix

Measuring the abundances of C- and O-bearing species in exoplanet atmospheres enables us to constrain the C/O ratio, which contains indications about the planet formation history. With a wavelength coverage going from 0.95 to 2.5 µm, the high-resolution (R ~ 70 000) spectropolarimeter SPIRou can detect spectral lines of major bearers of C and O in exoplanets. Here, we present our study of SPIRou transmission spectra of WASP-76 b acquired for the ATMOSPHERIX programme. We applied the publicly available data analysis pipeline developed within the ATMOSPHERIX Consortium, analysing the data using 1-D models created with the petitRADTRANS code, with and without a grey cloud deck. We report the detection of H2O and CO at a Doppler shift of around −6 km s−1, which is consistent with previous observations of the planet. Finding that a deep cloud deck is favoured, we measured a mass mixing ratio (MMR) log(H2O)MMR = −4.52 ± 0.77 and log(CO)MMR = −3.09 ± 1.05 consistent with a sub-solar metallicity to more than 1σ. We report 3σ upper limits for the abundances of C2H2, HCN, and OH. We estimated a C/O ratio of 0.94 ± 0.39 (~ 1.7 ± 0.7 × solar, with errors indicated corresponding to the 2σ values) for the limbs of WASP-76 b at the pressures probed by SPIRou. We used 1-D ATMO forward models to verify the validity of our estimation. Comparing them to our abundance estimations of H2O and CO, as well as our upper limits for C2H2, HCN and OH, we found that our results were consistent with a C/O ratio between 1 and 2 × solar, and hence with our C/O estimation. Finally, we found indications of asymmetry for both H2O and CO when investigating the dynamics of their signatures, pointing to a complex scenario possibly involving both a temperature difference between limbs and the presence of clouds being behind the asymmetry that this planet is best known for.

Enhancing Aviation Safety through AI-Driven Mental Health Management for Pilots and Air Traffic Controllers.

This article provides an overview of the mental health challenges faced by pilots and air traffic controllers (ATCs), whose stressful professional lives may negatively impact global flight safety and security. The adverse effects of mental health disorders on their flight performance pose a particular safety risk, especially in sudden unexpected startle situations. Therefore, the early detection, prediction and prevention of mental health deterioration in pilots and ATCs, particularly among those at high risk, are crucial to minimize potential air crash incidents caused by human factors. Recent research in artificial intelligence (AI) demonstrates the potential of machine and deep learning, edge and cloud computing, virtual reality and wearable multimodal physiological sensors for monitoring and predicting mental health disorders. Longitudinal monitoring and analysis of pilots' and ATCs physiological, cognitive and behavioral states could help predict individuals at risk of undisclosed or emerging mental health disorders. Utilizing AI tools and methodologies to identify and select these individuals for preventive mental health training and interventions could be a promising and effective approach to preventing potential air crash accidents attributed to human factors and related mental health problems. Based on these insights, the article advocates for the design of a multidisciplinary mental healthcare ecosystem in modern aviation using AI tools and technologies, to foster more efficient and effective mental health management, thereby enhancing flight safety and security standards. This proposed ecosystem requires the collaboration of multidisciplinary experts, including psychologists, neuroscientists, physiologists, psychiatrists, etc. to address these challenges in modern aviation.

Differentiable modeling for soil moisture retrieval by unifying deep neural networks and water cloud model

Machine learning has been widely used in high-spatial-resolution surface soil moisture (SSM) retrieval studies, but in recent years, this purely data-driven retrieval method has been controversial due to its lack of physical interpretability and generalization ability. Physical retrieval models based on the theory of radiative transfer equations respect physical laws, but their retrieval accuracy is usually affected by many insufficient accurate inputs, the complex model structure, and parameter adjustment method. In order to explore the retrieval method of unifying these two types of models, in this study, a differentiable model (DM) was constructed to realize the soil moisture retrieval at 10 m resolution based on Sentinel data. The differentiable soil moisture retrieval model takes the water cloud model (WCM) as the skeleton, and united the WCM and neural networks by implementing differentiable programming of the WCM in a machine learning platform. The differentiability makes the retrieval model trained by the gradient descent method the same as the neural network, which allows the retrieval model to be physical while obtaining more accurate retrieval results. Luan River Basin, Shandian River Basin, Maqu, and Lake Tahoe study areas with various land cover types and climate types were selected for model evaluation, and the performances of DM were close to that of the random forest model with Pearson correlation coefficient (R) of 0.747, 0.853, 0.838 and 0.792 in four study areas, respectively. While in the assessment of extrapolation capability of retrieval models, the DM showed its strong generalization ability and retrieval performance that exceeded that of the other retrieval models, with R of 0.786, unbiased root mean square error (ubRMSE) of 5.523 vol% and bias of 0.054 vol%. The DM synthesizes the advantages of both physical and machine learning models while providing high-resolution SSM estimates with acceptable accuracy for the study areas. This study creates favorable conditions for the realization of large-scale soil moisture retrieval with high resolution and high accuracy, and provides new ideas for the combination of machine learning and physical knowledge in other retrieval studies.

Shallow cumulus cloud fields are optically thicker when they are more clustered

AbstractShallow trade cumuli over subtropical oceans are a persistent source of uncertainty in climate projections. Mesoscale organization of trade cumulus clouds has been shown to influence their cloud radiative effect (CRE) through cloud cover. We investigate whether organization can explain CRE variability independently of cloud‐cover variability. By analyzing satellite observations and high‐resolution simulations, we show that more clustered cloud fields feature geometrically thicker clouds with larger domain‐averaged liquid water paths, smaller cloud droplets, and consequently larger cloud optical depths. The relationships between these variables are shaped by the mixture of deep cloud cores and shallower interstitial clouds or anvils that characterize cloud organization. Eliminating cloud‐cover effects, more clustered clouds reflect up to 20 W/m more instantaneous shortwave radiation back to space.

Noise4Denoise: Leveraging noise for unsupervised point cloud denoising

Existing deep learning-based point cloud denoising methods are generally trained in a supervised manner that requires clean data as ground-truth labels. However, in practice, it is not always feasible to obtain clean point clouds. In this paper, we introduce a novel unsupervised point cloud denoising method that eliminates the need to use clean point clouds as groundtruth labels during training. We demonstrate that it is feasible for neural networks to only take noisy point clouds as input, and learn to approximate and restore their clean versions. In particular, we generate two noise levels for the original point clouds, requiring the second noise level to be twice the amount of the first noise level. With this, we can deduce the relationship between the displacement information that recovers the clean surfaces across the two levels of noise, and thus learn the displacement of each noisy point in order to recover the corresponding clean point. Comprehensive experiments demonstrate that our method achieves outstanding denoising results across various datasets with synthetic and real-world noise, obtaining better performance than previous unsupervised methods and competitive performance to current supervised methods.

Intense formation of secondary ultrafine particles from Amazonian vegetation fires and their invigoration of deep clouds and precipitation

Intense formation of secondary ultrafine particles from Amazonian vegetation fires and their invigoration of deep clouds and precipitation

Tropical Deep Convection, Cloud Feedbacks and Climate Sensitivity

AbstractThis paper is concerned with how the diabatically-forced overturning circulations of the atmosphere, established by the deep convection within the tropical trough zone (TTZ), first introduced by Riehl and (Malkus) Simpson, in Contr Atmos Phys 52:287–305 (1979), fundamentally shape the distributions of tropical and subtropical cloudiness and the changes to cloudiness as Earth warms. The study first draws on an analysis of a range of observations to understand the connections between the energetics of the TTZ, convection and clouds. These observations reveal a tight coupling of the two main components of the diabatic heating, the cloud component of radiative heating, shaped mostly by high clouds formed by deep convection, and the latent heating associated with the precipitation. Interannual variability of the TTZ reveals a marked variation that connects the depth of the tropical troposphere, the depth of convection, the thickness of high clouds and the TOA radiative imbalance. The study examines connections between this convective zone and cloud changes further afield in the context of CMIP6 model experiments of climate warming. The warming realized in the CMIP6 SSP5-8.5 scenario multi-model experiments, for example, produces an enhanced Hadley circulation with increased heating in the zone of tropical deep convection and increased radiative cooling and subsidence in the subtropical regions. This impacts low cloud changes and in turn the model warming response through low cloud feedbacks. The pattern of warming produced by models, also influenced by convection in the tropical region, has a profound influence on the projected global warming.

Automated BIM-to-scan point cloud semantic segmentation using a domain adaptation network with hybrid attention and whitening (DawNet)

Deep learning-based point cloud semantic segmentation facilitates scene understanding and BIM modelling, but its success requires vast amount of labelled point clouds, which is laborious and time-consuming. To reduce point cloud annotation cost, researchers attempt to leverage synthetic point clouds, but the domain gap between synthetic and real point clouds deteriorates the segmentation accuracy. To address this issue, this study develops a BIM-to-Scan point cloud semantic segmentation approach to mitigate the domain gap between BIM and real point clouds, improving the segmentation performance on real point clouds. To this end, this study starts by proposing a BIM-based point cloud generation method, which uses FME and BIM models to automatically generate and label synthetic point clouds, decreasing the annotation cost. To fill the domain gap, a DawNet is invented by integrating a domain adaptation network with ZCA whitening operation and hybrid attention mechanism. Specifically, ResPointNet++ is used to extract geometric features and execute the segmentation task, which is then combined with a domain discriminator to perform domain adversarial learning, aligning the data distribution of BIM and real point clouds. To improve the performance of the DawNet, a residual learning block with whitening and a hybrid attention module are designed. These two modules help extract and exploit domain-invariant features to boost the generalisation and segmentation performance of the DawNet. Finally comprehensive experiments show that the proposed BIM-based method spends 0.5 person-hours to generate 0.45 billion labelled BIM point clouds, and that the developed DawNet achieves 17% and 11% more mIoU than ResPointNet++ and DANN. The ablation study also confirms the effectiveness of the hybrid attention module and ZCA whitening operation.

Semi-automated approach towards efficient HD Maps generation and verification with Lanelet2 formats

Abstract. HD Maps (High-Definition Maps) serve as crucial resources for the domain of autonomous vehicle. Because HD Maps can provide detailed and accurate road information, the generation of HD Maps has been a labour-intensive and high cost. This research presents an innovative and semi-automated approach for efficient HD Maps generation by using assure mapping tool with deep learning techniques and mobile laser scanned point cloud geometry. The proposed method starts with data collection from various sources such as images, LiDAR point clouds, and integrated INS/GNSS trajectory data. These data are labelled by using a pre-trained model. After finishing post-labelling, these data are subjected to deep learning training by using VoxelNet and Yolact++ framework and leading to the generation of an AI model. The tool effectively recognizes and categorizes features such as road surface markings, traffic signs, and traffic lights, which can be further expanded as per requirements. Finally, the output format can be converted to OpenDRIVE, Lanelet2, and other else. Hence, the extracted lane lines can compare to the manual mapping data for verifying the accuracy. This study demonstrates that the proposed approach can be instrumental in streamlining the HD Maps generation procedure, reducing manual labour, and enhancing efficiency. The assure mapping tool proves to be an effective instrument, particularly when powered by deep learning algorithms and point cloud geometries, in the creation of reliable, comprehensive, and application-ready HD Maps.

Simulating the seeder–feeder impacts on cloud ice and precipitation over the Alps

Abstract. The ice phase impacts many cloud properties as well as cloud lifetime. Ice particles that sediment into a lower cloud from an upper cloud (external seeder–feeder process) or into the mixed-phase region of a deep cloud from cirrus levels (internal seeder–feeder process) can influence the ice phase of the lower cloud, amplify cloud glaciation and enhance surface precipitation. Recently, numerical weather prediction modeling studies have aimed at representing the ice crystal number concentration in mixed-phase clouds more accurately by including secondary ice formation processes. The increase in the ice crystal number concentration can impact the number of ice particles that sediment into the lower cloud and alter its composition and precipitation formation. In the Swiss Alps, the orography permits the formation of orographic clouds, making it ideal for studying the occurrence of multi-layered clouds and the seeder–feeder process. We present results from a case study on 18 May 2016, showing the occurrence frequency of multi-layered clouds and the seeder–feeder process. About half of all observed clouds were categorized as multi-layered, and the external seeder–feeder process occurred in 10 % of these clouds. Between cloud layers, ≈60 % of the ice particle mass was lost due to sublimation or melting. The external seeder–feeder process was found to be more important than the internal seeder–feeder process with regard to the impact on precipitation. In the case where the external seeder–feeder process was inhibited, the average surface precipitation and riming rate over the domain were both reduced by 8.5 % and 3.9 %, respectively. When ice–graupel collisions were allowed, further large reductions were seen in the liquid water fraction and riming rate. Inhibiting the internal seeder–feeder process enhanced the liquid water fraction by 6 % compared to a reduction of 5.8 % in the cloud condensate, therefore pointing towards the de-amplification in cloud glaciation and a reduction in surface precipitation. Adding to the observational evidence of frequent seeder–feeder situations, at least over Switzerland, our study highlights the extensive influence of sedimenting ice particles on the properties of feeder clouds as well as on precipitation formation.